Cluster Locks health check fails in Jira Server
Overview
In JIRA Data Centre some operations require the a lock to be placed across all nodes for operations that should only be operated on one node at a time, for example taking a backup of the index. This Health Check will check the age of the cluster locks and report a warning in the event that a cluster lock is found that is older than five minutes.
Understanding the Results
| Icon | Result | What this means |
|---|---|---|
 | The health check passed successfully. | No lock was detected that is older than five minutes. |
 | Node '<node-name>' has been holding cluster lock, '<lock>', for <seconds> seconds. | A lock has been held for longer than five minutes. |
| Problem | Suggestion |
|---|---|
| A Node has not released a lock. | Check that the node reported in the health check is responsive and running. Review the logs of the node to see if there are any errors or exceptions. In the event that the node is not running, or unresponsive, a restart of the affected node is expected to clear the lock. Prior to doing so it is recommended to collect thread dumps as per Generating a Thread Dump. |
| Database is timing out when completing work. | Check the performance of your database. This can be done by:
|
| A task is taking a long time to complete. | Follow our Troubleshoot performance issues in Jira server KB. An LDAP communication issue can also be causing this: Unlimited LDAP read timeout can cause Cluster Locks health check to fail if there are communication issues |
Providing Information to Support
In case you are unable to troubleshoot and fix the problem by yourself, please create a support ticket at support.atlassian.com and attach the following information to the ticket:
- Take a Screenshot of the Health Check results.
- Collect a Support ZIP from each of the Data Centre nodes.
- Any collected information from the suggestions in this document.